armaab/hyperlpr-train

Issues

• 能更新一下HyperLPR中用GRU的训练网络吗

  #5 opened 6 years ago by deepblacksky
  20

• 识别训练生成多个.h5 .pb文件这个要怎么转成HyperLPR库所能使用的*.caffemodel和*.prototxt文件呢？

  #20 opened 4 years ago by azhon
  0

• https://gitbook.cn/gitchat/activity/5c9c902ddd42c631eb85beed

  #17 opened 5 years ago by hoolang
  1

• 模型训练完后，保存的h5文件载入hyperLPR中无法识别训练样本？

  #6 opened 6 years ago by Pau1code
  6

• 训练出来的模型文件应该代替原先的哪一个文件？

  #15 opened 6 years ago by payphone131
  1

• test_nn.py运行结果错误

  #19 opened 5 years ago by a790399215
  4

• 用训练出来的模型文件，车牌识别不出来

  #14 opened 6 years ago by xiasys001
  2

• 我愿意维护

  #12 opened 5 years ago by baiyoulong
  4

• 请教一下，运行训练脚本一直显示loss:nan是怎么回事？

  #9 opened 6 years ago by axltfytx
  7

• 请问一下，训练完成的模型在load_model时怎么正确加载ctc自定义损失函数，可以提供一个示例吗

  #18 opened 5 years ago by Soulempty
  0

• cnn+ctc loss

  #13 opened 5 years ago by MisayaZ
  1

• Please update code with GRU, thanks!

  #16 opened 5 years ago by soufianesabiri
  0

• Can you give an example of predicting code with the trained h5 model?

  #8 opened 6 years ago by 1474123989
  1

• 请问这个训练大概使用多少张图片来训练比较好?

  #10 opened 6 years ago by linzhi123
  1

• 如果在训练过程中,中断了训练,下次可以跳到断点开始训练的吗?

  #11 opened 6 years ago by linzhi123
  1

• why NUM_CHARS+1

  #7 opened 6 years ago by 1474123989
  1

• 关于export()函数

  #4 opened 6 years ago by seekFire
  1

• 请问网络的结构是什么样的?

  #3 opened 6 years ago by ceceha
  6

• the trained model can not recognition the Chinese characters

  #2 opened 6 years ago by cuicuizhang1989
  2